 Mr. Fahy, you're on the two. Yeah, I'll go ahead. Hi Audrey, it was great hearing from you. My name is Mustafa, I'm from Pakistan. I have a few years of work experience working with the governments, particularly the health department in Pakistan in Punjab, the province Punjab. So I'm very impressed by how you guys tackled the COVID breakout. And like Stefan said, in Pakistan it was mostly from a top-down approach, where a monitoring control center was set in the federal level run by doctors and the army. And then based on that, different interventions were implemented at the province level, using the already some of the structure of the health department and some of the resources of the army. But I'm really curious about if you could talk about in some more detail about how you use this, have you combated the COVID pandemic through IT? I would like to hear that in more detail. Yeah, certainly. Of course, we have published like a lot of papers in Sudan. I think our Ministry of Foreign Affairs even had a dedicated website called the Taiwan Model of Combating COVID-19 that lists a lot of like very fine details. So I'll just case my MOFA and Health and Welfare links respectively to here for further reading. But to summarize, I think what we have done is trusting the citizens. Instead of ordering citizens what to do, we trust the people on the front line, including health workers and social workers, to understand the situations better than us. So we focus on two things. First, getting the collective intelligence's input in a coherent way to find the common grounds of innovation, despite the very different lived experiences in urban and rural areas and so on. So this is the first thing. And second, after we got a genuinely a good idea, a mask rationing map, a contact tracing stuff and things like that, we work on amplifying it as quickly as possible using whatever apparatus that there is in preferably 24 hours after we discovered this genuinely good idea. This is so we can pivot. When we first introduce an idea, of course, it's going to have bias. It's going to have unforeseen consequences and so on. But if we do this iteration quickly enough in the CCC, we admit our mistakes very quickly in 24 hours and the good ideas they are implemented at most next Thursday. So we do deployment every Thursday. So people learn this cadence very quickly. That their input actually matters. It could just be leading to the fashion brands adopting pink as the color of the day or whatever their new good ideas are. And so we treasure people's actual experiences and work on ways to converge them into coherent blended volitions. And so, of course, we leverage existing infrastructures as I alluded to, like our universal health care IC card, which is very important. We leverage the fact that the pharmacies and later on convenience store all have secure lines into the administration of health and welfare like I think 90% or more have fiber optic connections and so on. But that help on the real timeness of data because otherwise it will not be trustworthy. We leverage the kiosk ATM like machines in convenience stores. We repurpose them so they can read those health IC cards and so on. So we change a lot of IT systems and all of them like SMS and QR code exist before the pandemic. So people do not have to learn a new set of norms. This always technologies adopt adapting to where people are never asking the people to adapt to where the technologies are. And so like my grandma almost 90 years old was my focus group, one of my focus groups. And she recommended her younger friends 77 years old. And we always make sure that they comprehend these new measures and they also offer very good feedback. Like we're about to implement go home and idea of pre-registration and broader distribution convenience store. Well, we initially had this idea of people can just use the academic card on the ATM to transfer a couple of cents to prove that they are who they are. And then use the receipt of transferral to redeem the mask that they pre-order, which sounds very good on paper. But it doesn't work. Grandma Yang 77 years old said that she always use paper notes to withdraw to wire money out. She only use the ATM to withdraw cash because she want to know that she after a typo doesn't transfer her savings out. She feels insecure. She would instead instead of her debit card want to use her health card because it's not linked to her bank account. And she knows that she will lose nothing. And she is okay to pay through coins on the counter to the convenience store on the kind of pre-registration fee and so on. So it sounds less efficient, but it's actually the only way to get the elderlies on board. And once we probably credit it, I said very probably it's grandma Yang's idea. Then the 77 years old and the 66 years old become key opinion leaders. And they teach each other that they can actually now go to convenience store and use their health card. It's just one anecdote, but I hope it conveyed the feeling. Thank you. So Audrey, one key element that we discussed, which was that to a large extent data stewardship is always also making sure that actually we not just go from data to insight, but from insight to action. And so you tell us a little bit about this, because you said it's like a fast kind of decision cycle, right? So you have new insights that give you a different perspective on reality and as a result you act differently. So how did you establish that? And what does it take to have that kind of decision, rapid decision cycle based upon new input? Yeah, so this is something that we always needed to do before the pandemic or something more severe happens, right? Because you can't really ask people to suddenly trust the government, but you can as a government officials trust the people before a disaster. And if you build this kind of trust enough, then some people eventually trust back, which is the underpinning for this kind of co-creation to work to turn their insights into action. So before the pandemic, since I become the digital minister in 2016, for four years before the pandemic, I've published all the meetings that I have with lobbyists and journalists and even internal cross-ministerial meetings that I chair into this say it website. So anyone can see how digital minister work basically. And this conversation with your permission, of course, I will also publish the transcript of our conversations, but you can prefer if you want to be identified only as audience member, right? So this is not about violating your privacy. This is about making sure that people understand who I'm talking with and what content I'm talking with so that people understand how of policymaking, the why of policymaking and never just the what of policies. This is the most important thing. And once people get into the habit of seeing me talking about emerging technologies and things like that, then we can deploy more advanced platforms so that people can comment on those ideas. Now, when people comment on those ideas, the main worry of public service is that it would just be noise, right? People are used to things like Facebook, which is the digital equivalent of a nightclub of very loud music. You have to shout to get her addictive drinks, private bouncers that escort you out. And so mental health hazard to young people and also adults, let's be frank. So basically, it's like the nightlife district. Now, don't get me wrong. I respect the entertainment sector, but I don't think we should have town hall discussions in our local nightclubs period. So what we need is this is my office, by the way, this is the social innovation lab. Things that are genuinely public infrastructure, like a public park, which is my office, how it's modeled after, or like PTT, the campus that I mentioned at the beginning and so on. So then we can hold on to the discussions. This is Polis, a free software, free as in freedom for assistive intelligence powered conversation that is the opposite of anti-social corners of social media. This is the pro-social corner of social media where people compete to find consensus, not conspiracy theories. So in 2015, we started deploying that on national level conversations. And nowadays, it's very typically a public infrastructure because it's Polis GOVTW, no longer the GZOVTW, no longer just in a free software community, but just in the civil service. So any civil servant can start new Polis conversations, just like you can start a Google survey. It becomes an everyday thing. So this survey is weekly made. People contribute their own survey items. So in 2015, we talked about the Uber case, for example. But we talked about something very specific. These are the facts around people driving to work, picking up random strangers that they met through an app, charging them for it, but they don't have a professional driver license. How do you feel about that? So we don't talk about ideologies like geek economy versus sharing economy or whatever. But we talk about this practice, okay? And then we ask people for three weeks, how do you feel about the same folks? Who may feel happy, they may feel angry. That's okay. But after we set up this pro-social conversation, always after three weeks, they converge on real insights based on people's actual feeling and experience. That then we take those insights and then we make them into decisions. This is the focus conversation method, the four step method or ORID method that we use. Now, the user experience, the human experiences like this. You see a fellow citizen, they have a sentiment and you may agree. If you do, your avatar move toward me. If you don't, your avatar move away from me. So upvoting and downvoting is portrayed as social distancing, right? But it's not anything measured by headcount. So even if you mobilize 5,000 people to vote exactly as you do, it's just one dot here. It doesn't increase the area of the K-Means cluster. And therefore, it doesn't affect the outcome at all. So trolling doesn't pay. And after all, there's no reply button. So you can't even troll me. The only thing you do is to find some other feelings that convince people across the aisle because we attribute more decision-making power to anything that convince everyone. The more polarized you are in the K-Means cluster, if you find something that you and this diagonal person agree, then it counts more toward our shared agenda. And we have a real-time visualization for that. So people agree to disagree on just a few ideological things like whether it's geek or sharing economy. But people almost all agree with their neighbors on most of the things, most of the time, that insurance is important. Not undercut existing meters is important. Empowering local church and temples to form their own fleet to serve the place where taxes don't serve. That's important. Search pricing, very important, and so on and so forth. So when everyone, Uber passengers and taxi passengers and their drivers all agree on these things, we then take this to a multi-stakeholder meeting live streamed and ask Uber and everyone, okay, here is the common feeling of the people, why are you not informing to the norm? So I call this the people-public-private partnership. The people sets the norm. The public amplifies the norm. The private sector just implements the norm. And this is obviously a success because for quite a few years now, Uber is a legal taxi company, the Q-taxi. But we also have local co-op like Donkey Move that works as a social entrepreneur to serve the underserved areas where Uber wouldn't go into. And they share the same multipurpose taxi act that is co-created by the people. So again, this is something that elays people's fear, uncertainty, and doubt and turn their energy of tension into an energy of co-creation thanks to the real-time visualization that people understand that will lead to decisive action. And we do that all before the pandemic. So it counts during the pandemic. Estefan. Hi, Minister Tang. Thank you for your sharing this morning. I like your comparison of social media to a nightclub and also how it would be interesting to see your tips on how to facilitate the focus group with your grandmother and with your grandma in it. But my question really was on your, my friend, on I2-2 outline. So I was wondering on the back end, say, people can call just to share about anything. Yes. They have healings or anxiety or injuries about COVID-19. So on the back end, like how do you, the call center, how does it categorize the data? Yeah. How do you put it into notes, for example, and then how is it consolidated and activated so that it informs the decision-making to share it. So I'm interested to see how it is consolidated, how it is processed and then disseminated into relevant decision-making. Yeah. The I2-2 back end is the CHT, our largest telecom. So they already have very good customer feedback call center systems setting up with the elastic workforce that, as I mentioned, can invite the social sector to charities. So GR, our largest charity on disaster recovery, they do their work internationally as well. They have a lot of very, they're also a Buddhist organization, right? So they have a lot of compassion, empathy, and so on, just as part of their daily training. So these two factors together is necessary. It's necessary to have a call center that can identify the shared topics that the people were talking about and escalate whatever that's not in the Q&A. But it's equally important or not necessarily Buddhist, but people who train the art of compassion to actually draw out the authentic feelings, not just the rushed interpretations, the rushed demands and things like that. Because we know that there's demands. We know that people want more masks. They want a vaccine sooner. But what we want is that, what's the authentic experience that led to this feeling of shortage? And what we want is the insight. So if you've done experience interview in a service design journey, you know what I'm talking about. What we want is to put people's authentic experience on a map so that we can count the dots and see that here is the pain point. We need to work on this touch point first. So it's both an art of human communication, but also a science of service design and both need to be together. But I think GovLab is expert on this. There's hundreds and hundreds of pages of how to build such techniques, social technical systems. So please refer to GovLab to how to set up this kind of teams. Thanks, Audrey. Any other question? I have a quick question also, Audrey. We're going to, I mean, this group focuses on also how do we not just get data equity, right? And how data is collected and used. But how do we also get question equity? Which, who actually determines what the questions are that are getting answered? And how do you go about that? In a more participatory way. And so wondering whether Audrey, you have also done some work, not just in generating insight, but actually generating questions. Yeah, definitely. And so you get to hear how, in any example, or how you've done that. Yeah, we have a entire system to do that called the participation offices and collaborative meetings and so on. And here are the two relevant links. The first is in English. The second should probably work with Google Translate. And the way it works, very simply put, is that each ministry of a dedicated team called participation offices that are in charge of citizen engagement. Now we meet every month and we vote on the two topics per month that we want to have a multi-ministrial conversation about. But it includes all the stakeholders and so on. A popular way to determine what goes on the ballot is to start a petition on joining the GOV, the TW, which after you collect 5,000 signatures, demands a ministry of response. But if whatever you're asking is cross-ministrial, for example, a 17-year-old a few years back asked that we ban plastic straws from the bubble tea, which is our national drink, takeouts gradually, right? But this is obviously not just the environmental protection administration. It's also the Ministry of Economic Affairs among other interior and so on. So we need to get everyone on board. And that is how the PO's participation offices work is that once they voted this in, we invite this person who's not yet an adult, but she has a gender setting. And whomever who wants to participate from the 5,000 signatures, we invite five from them. But the other 4,000, they can watch through livestream and participate through Slido. And then we map out the stakeholders and we discover how to kind of face out plastic straws gradually. And we turn that into action. And people can always go back to see the transcript and things like that in the collaborative meetings. Now, what we found is that we deliberately assign into the breakout groups to facilitate us the unrelated ministries. So when we talk about the plastic straw, maybe the unrelated ministry like the, I don't know, Coast Guard, host the conversation. Because Coast Guard is on the business end, right? They deal with the result of the plastic straws, but they're not directly involved in the policy making stage. Or when we talk about tax reform, how to fire our income tax, which was according to the petitioner and I quote, explosively hostile to Linux and Mac users. And of quote, we invite not the tax agency, but maybe the ministry of culture to hold a breakout group. But when we talk about after the ocean policy fishing and so on, then we invite the tax and culture agency people as the breakout group. So you get a feeling basically those people, they also serve or fish in their spare time. The Coast Guard, again, maybe they buy bubble drink, the ministry of culture person, they file their tax themselves. So in their daily life, they're also citizens negatively affected by bad policies. It's not that just because they're civil service, they have to defend everything, right? For unrelated ministries, they are every bit as mad as other petitioners. Now, for the petitioners who actually join and people watch online, the energy is someone with deep knowledge of how civil service work somehow advocating for their cause. And they never seen this before, right? Their counter expertise doesn't have public administration experience. The public administration always defend whatever existing policy is, but the restoration officer is a rare blend of the two. And so we can find workable solutions on those questions as raised by the popular vote of the petitions almost always. So sometimes the solution exists on the civil service. Sometimes it resides on the social sector. Sometimes it means we need to pressure Facebook together and threaten social sanction. But whomever is accountable for the change, it's very clear that a civil service are also citizens. And this is how the system should work. Great. Any other questions from the cohort here? Yes, please, please. Yeah, I have to know me. I'm not used to Google. I have no idea on how to navigate or even sure. Just unmute and speak. This should be a call to action. Unmute yourself. Thank you. Yes. Thank you. A very important question. As a matter of fact, the indigenous people in Taiwan also have a shorter lifespan. I don't have the exact number, maybe six, seven years, compared to the average of population. As a result, when we started vaccination, they get a 10 year kind of discount, right? So even if they're younger, they qualify for the older age group by 10 years. So that's the first thing. And the second thing is that we ensure because the universal healthcare is at no cost to them anyway. We bring the vaccination to them. Everyone else need to pre-register and go to the vaccination spot. But for indigenous people, nations, we do the reverse. So we just bring the entire vaccination in a kind of mobile delivery fashion and bring to them. We later on would also do that for the very, very old people. But I think the indigenous nations receive that mobile vaccination first if I'm not remembering wrong. So that's the second plan. But at the end of the day, I think this inequity is about a cultural inequity. A lot of our vaccination or masquerading or whatever information I must admit is only bilingual. It's available in Mandarin and in English, which is strictly speaking wrong because we have 20 national languages here and 16 of which are indigenous. So by law, the National Language Act, we should provide at least machine translation or interpreter to the 16 languages. The fact is that we do not yet have sufficient natural language processing beta to produce reliable translation that we are sure that will not inadvertently harm people. But we are working on that on the so-called low resource languages using transfer learning. For example, Amis have a lot of people. Sakilaya is somehow related to Amis linguistically very important to know this. Different nations linguistically. We must now work with the experts to do transfer learning so that machine learning that works here will also work here. This is one of our most important investments and he qualifies as infrastructure money, even though it's not made out of concrete, but just of bits in the next couple of years and so on. And I think when culturally, they feel that they speak in their kind of native language about how vaccines work and can freely remix the cute dog memes or whatever. I think that will also lead to more kind of self-awareness of how as a nation, we're going to negotiate our vaccination with the central governments semi-diplomatically. That is the end goal. But language and cultural equity in addition to the education and broadband equity that we already have, I think need to happen before that happens. Hope to answer your question. Is it in the kind of, what do you mean you refer to that is that if you have a cultural language barrier, and you also have a barrier, but how to display out in the collective intelligence, for instance. Well, and is there need for, there's a lot of discussion about data sovereignty, for instance, or indigenous data sovereignty. Is that something that comes up also in Taiwan? Yeah, good question. I think the Lan Yu, the Orchid Island people try to issue their local money using Ethereum. But Ethereum at that time doesn't have the soulbound tokens which I'm working on with Vitalik going to interview him tomorrow on this. So without a native on-chain way to represent the indigenous idea of meaningful participation to the community, or I don't know how to call it in the Orchid Island language. After all, I don't know all 16 languages. Anyway, but for the Ataia, you'll be in Gaia. So it's like the South African-African idea Ubuntu, right? So it's not that groups are assemblies of individuals. It's the other way around. One becoming an individual as a meaningful intersection of our contribution to the various groups. So that's a fundamental insight. And that's not there in Ethereum or Bitcoin community at the time that Orchid Islanders did their local coins, right? So at the end of the day, it become like people who buy the wallets, you know, buy the wallets. Later on, we will see this dynamic on NFT, which may not be fungible, but it's transferable. So anyone with the most money just win the NFT game and it become just serving the need of showing your status across lockdowns, which is, I guess, something that's worth paying for for some people. It's as far from indigenous communal spirit as you can get. So it's not a good fit is what I'm saying. So what we're trying to do is to find again, asking the technology to fit the society, not adapting the society to fit how Bitcoin work or how Ethereum work or how national identity cards work, which is ultimately about individual sovereignty, which is not a good map to the culture. So, so we really want something like so bounds tokens or nowadays we call it decentralized society desox to actually work. And when that works, I think we'll have much better social substrate on top of which to grow like the dawn of civilization style, new ways of social arrangements and so on. But we first need to upgrade the Internet from a just Internet of things and Internet protocol addresses into a true Internet of beings and we're not done that yet. So we need to focus on that in the next couple years. And again, we'll see to it. That's fine. Thank you. Thank you, Minister. I really like your presentation there. Internet of things. Yeah, it includes the dog. So the dog agrees to. Yeah, so I was wondering minister does the old. Some of these things in the sets for cross sector. And then trying to address could be addressed. You see some. Yeah, so, sorry, I didn't get the last sentence. It could be what you see the data sets which used to be across sectors and then. Yeah, yes. Okay. Thank you. Yeah. So, what the ministry of digital affairs will help do is to organize to design and run the presidential hackathon, which is the institution support for the kind of cross sectoral data. We call it coalitions, but you can also get a collaborative. I think that's the term that God love uses. And so I just changed that to say data collaborative anyway. So the ministry of digital affairs proud organizer of next year's presidential hackathon is the kind of proponent of co-creation of this map, for example, which measure air quality PM 2.5, but you can't really tell which sector contributed which and many of this are in the education sector where the primary schoolers are learning about data stewardship based on the Arduino or raspberry stations that they put forward on their balcony or whatever that determines whether their parents walk to nearby market or have to go in a bus or a car because of the PM 2.5 pollution level. So they contribute meaningfully to their family, but on this area, which is the industrial park area, it is done by the municipal government because they own the lamps public lamps in those industrial areas and the teachers can't really break in and install So basically we work on the protocol. We work on the distributed ledger that holds everyone to account. We work with the National Center for high speed computation, which is part of ministry and soon council commission of national science and technology. So we work with an STC and moda together to ensure that the even primary schoolers have the computation power to test their alternative models of how those things work. And this is what we call data competence, not just literacy where they learn about the inferences of the models but competence, the freedom to fork the models so that they can build their own hypothesis and so on. And every year, we work with the presidential office to give five awards. That's just like this to the winners of presidential hackathon. And here is a micro projector. I don't know whether you can see it underneath the shape of Taiwan. And if you turn on the micro projector, it projects Dr. Tsai Ing-wen, our president, handing you the trophy. So this is very main type like this, this trophy literally describes itself. And then you see Dr. Tsai Ing-wen on that projection, promising whatever you did in the past three months locally will become national policy with the funding personnel and regulation required. So this is how we select presidential annual agenda partly with crowdsourcing. And the data collaboratives built this way of course need to prove its value to pretty much everyone. And the way that they surface themselves is to lobby for votes. Now I don't have time to go into details, but we use a new voting system called quadratic voting. So all the projects need to declare publicly which of the 169 are sustainable goals that they correspond to, or at least one of the 17 initially. So it must be a public good. And now we're all public goods, 200 different projects. Each citizen in Taiwan, actually everyone with a local SMS number, receive 99 points which you can allocate. And then it's, if you vote one vote to a project is one point cost, two votes, four points, three votes, nine points of quadratic was 99 points. You can only vote nine votes to your favorite pet project 81 and you have 18 left. So you will vote instead of squandering them for votes to something else. So learn more about is in the process you still have two left so vote for right so on average people vote for four or more projects maybe they'll do a seven and seven or whatever. By the end of the day we have the statistics that we want so that when we choose the top 20 to incubate not only everyone feel they have won because they've supported probably one of the 20 but also the other 200. You also they have won because they can just reallocate their energy to the top 20 that has the most synergy with them because the quadratic voting patterns. And so everybody wins afterwards. So this is how new mechanism design can turn a zero or negative some game of voting internet voting on your popular pet project used to be the most divisive of things into something that's genuinely co-created and finally leads to decisive policy. And if you want to learn more just type presidential hackathon very easy to remember branding. Thank you for your question. Hope to answer your question. And yeah, there's always an international version of that as well. Yeah, yes, yes, yes on the net zero in practice. So instead of all 17 the international track focused on the only thing of the 17 that is equally felt because carbon dioxide and other greenhouse gases doesn't know jurisdictional boundaries and like the other SDG goals. So we focus on the 13 for our international track. Wonderful. Well, I'm sensitive to your time with me you're a busy person. So you've given us an hour, which was a privilege. And you've given us links here as well that we will follow up another way that people can follow recent developments working out. Well, my Twitter, my Twitter and maybe like literally follow. I think it's still in a nine life district, but not as loud as. Thank you. Thank you. Thank you. Thank you. Thank you very much.